The accuracy and stability of motion prediction are crucial for the safe planning of autonomous driving systems. The widely used attention mechanisms effectively improve prediction accuracy. However, their computational cost grows quadratically with sequence length, presenting challenges for handling complex, large-scale scenarios. The attention patterns of motion prediction tasks exhibit significant data-related sparsity, indicating that not all scene elements are worth interaction. Therefore, this paper proposes a novel motion prediction framework that enhances the efficiency and stability of interaction fusion while achieving promising prediction accuracy. Firstly, an adaptive self-distillation attention module based on sparse interaction graph is designed. This module adaptively filters high-value sequences for each target to achieve sparse attention calculation and balance scene scale. Due to the lack of interaction labels in the original dataset, a self-distillation strategy is employed for model training. Secondly, a multi-stage dynamic anchor decoder is introduced that leverages the information filtering and aggregation capabilities of the sparse attention mechanism to improve prediction accuracy. The decoder dynamically updates the target anchor representations as the prediction progresses, ensuring consistency between interaction states and the prediction process in long-term forecasting. This approach effectively focuses attention calculation on the most relevant scene context fusion and trajectory decoding at each prediction stage. Validation results on Argoverse 1 and Argoverse 2 demonstrate that the proposed method achieves competitive accuracy while effectively reducing computational resource consumption. The proposed method can also serve as a plug-in that can be seamlessly incorporated into standard motion prediction pipelines to optimize scene interaction, making it more friendly to low-cost devices.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Self-Distillation Attention for Efficient and Accurate Motion Prediction in Autonomous Driving


    Contributors:
    Lu, Ziheng (author) / Cai, Yingfeng (author) / Sun, Xiaoqiang (author) / Wang, Hai (author) / Lian, Yubo (author) / Chen, Long (author)


    Publication date :

    2025-05-01


    Size :

    2884548 byte




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    English




    Efficient Baselines for Motion Prediction in Autonomous Driving

    Gomez-Huelamo, Carlos / Conde, Marcos V. / Barea, Rafael et al. | IEEE | 2024


    Efficient Semantic Segmentation via Self-Attention and Self-Distillation

    An, Shumin / Liao, Qingmin / Lu, Zongqing et al. | IEEE | 2022


    System and Method for Motion Prediction in Autonomous Driving

    SULLIVAN ALAN / CHEN SIHENG / WANG JUN et al. | European Patent Office | 2023

    Free access

    AI-TP: Attention-Based Interaction-Aware Trajectory Prediction for Autonomous Driving

    Zhang, Kunpeng / Zhao, Liang / Dong, Chengxiang et al. | IEEE | 2023